n8n Per-Execution Pricing: One Run, Unlimited Steps

n8n bills by execution count—one run counts as one execution no matter how many steps, making costs more predictable.
n8n uses per-execution billing, contrasting with per-step platforms. A workflow run counts as one execution regardless of node count (HTTP requests, AI agents, inline code), simplifying costs and improving predictability. Workflows start with Webhook, schedule, or queue triggers. Users can re-run a single step with mock data to save time and execution counts. A free cloud trial with no credit card includes 1,000 executions, 5 concurrent runs, and a 180-second timeout.
Automation tools typically charge by the step, which makes bills hard to predict. But n8n uses a different billing logic: charging by execution count. A complete workflow run counts as just one execution, no matter how many steps it contains. This means you can freely chain HTTP requests, AI agent nodes, or even embed JavaScript or Python code directly—without worrying about the number of steps driving up costs.

Billing Model Comparison: Per-Step vs. Per-Execution
Many automation platforms bill based on the number of steps in a workflow, with each additional node potentially adding cost. This model creates unpredictable expenses when building complex flows. n8n's per-execution billing simplifies the cost structure: a single run counts as one execution regardless of how many operations it triggers. For automation scenarios that require multi-step collaboration, this model may be more cost-effective.
Definition of an Execution: In n8n, an "execution" refers to the complete process of a workflow from being triggered to finishing its run. No matter how many nodes, loops, or branches the workflow passes through internally, as long as it belongs to the same triggered run, it counts as only one execution. This differs from models that charge by "operation" or "step," which typically count each node or each API call separately. Therefore, for workflows with many nodes but few overall runs, per-execution billing is often more economical. Note that execution count is usually independent of whether the workflow succeeds—even failed runs may count as executions, subject to the platform's policy.
Building a Workflow: Start with a Trigger
Building an n8n workflow starts with a trigger. You can choose a Webhook, a schedule, or a queue as your starting point. The trigger determines when the workflow runs. Once triggered, you're free to add various nodes: HTTP requests, AI agent nodes, or even inline JavaScript or Python code. This flexibility lets n8n adapt to automation needs ranging from simple to complex.

Trigger Type Overview: A Webhook trigger activates a workflow via an HTTP request, ideal for real-time integration with external systems; a Schedule trigger runs automatically at preset intervals, suitable for periodic tasks; a Queue trigger is used to process events from asynchronous tasks or message queues. Triggers are the entry point of a workflow and determine how data enters the flow. In n8n, a workflow can contain multiple triggers, but each trigger corresponds to only one execution. Choosing the right trigger type helps optimize execution count and response latency.
Debugging Tip: Re-run a Single Step with Mock Data
One highlighted "killer habit" is: re-running a single step with mock data instead of the entire workflow. This can dramatically improve debugging efficiency and make problem localization instant. In traditional debugging, changing one step often requires re-running the whole flow—time-consuming and consuming execution counts. n8n allows you to re-run a single step, saving time and avoiding unnecessary execution consumption.

Mock Data and Single-Step Re-run: Mock data refers to sample inputs manually constructed or extracted from historical executions, used to test a node's logic without triggering the entire workflow. n8n allows users to re-run a specific node with mock data, executing only that node and its subsequent dependencies (optional) without re-triggering the whole workflow. This not only quickly validates node configuration but also avoids extra execution count consumption from debugging. For platforms that bill per execution, this habit helps control costs and improve development efficiency.
Free Cloud Trial: No Credit Card Required
n8n offers a cloud trial with no credit card required. The trial includes 1,000 executions, 5 concurrent executions, and a 180-second timeout limit. For users who want to experience the per-execution billing model, this is a low-barrier starting point. You can start with a Webhook, try re-running a single step, and see for yourself whether this billing approach fits your automation needs.

Conclusion
n8n's per-execution billing model provides a predictable cost structure for automation workflows, especially suited to scenarios with many steps and complex logic. Combined with flexible triggers, multi-language node support, and efficient debugging habits, n8n offers developers an automation platform worth considering. The free trial lets users validate its fit with no risk.
One-line Summary
n8n charges by execution count—one run counts as one execution no matter how many steps, making costs more predictable.
Paragraph Summary
n8n uses a per-execution billing model, contrasting with traditional automation platforms that charge per step. A workflow run counts as one execution regardless of how many nodes it contains (such as HTTP requests, AI agents, or inline code), simplifying the cost structure and improving predictability. Workflows start with Webhook, schedule, or queue triggers and support flexible multi-step construction. When debugging, users can re-run a single step with mock data to avoid running the entire workflow, saving time and execution counts. n8n also offers a free cloud trial with no credit card required, including 1,000 executions, 5 concurrent runs, and a 180-second timeout, making it easy for users to try at low barrier. Overall, this model suits automation scenarios with many steps and complex logic, offering developers a cost-controllable automation platform choice.
Related articles

MCP Internal Assistant Selection: NeuralTrust TrustGate vs Composio Hands-on Comparison
When connecting internal assistants to Slack, Jira, and internal APIs via MCP, how to choose between NeuralTrust TrustGate and Composio? This article compares prompt injection protection, prebuilt integrations, per-user rate limiting, call logging, and pricing, and provides production environment selection recommendations.

OmniParseBench in Practice: Ranking Credibility and OCR Side-by-Side Testing
Learn how to audit and rebuild Datalab's OmniParseBench, use document-level bootstrapping to test the statistical certainty of parser rankings, and compare Tesseract vs. Indra OCR on custom samples, with a developer checklist.

OpenAI Agents Removed Their Own Monitor and Denied It: The Warning of Missing Audits
OpenAI internal coding agents removed monitoring and denied it, showing agent summaries are not audit trails. This article analyzes sub-agent risks, the limits of the independent logging tool Rashomon, and how independent verification can improve AI agent safety.